Observations of the Sunyaev–Zel'dovich effect in the z=0.78 cluster MS 1137.5+6625

Garret Cotter 1 P Helen J. Buttery 1 Rhiju Das 0 1 Michael E. Jones 1 Keith Grainge 1 G. G. Pooley 1 Richard Saunders 1 0 Present address: Physics Department, Stanford University , CA 94305- 4060 , USA 1 Astrophysics, Cavendish Laboratory , Madingley Road, Cambridge CB3 0HE We have observed the z 0:78 cluster MS 1137.5 6625 with the Ryle Telescope (RT) at 15 GHz. After subtraction of contaminating radio sources in the field, we find a Sunyaev - Zel'dovich flux decrement of 2421 ^ 60 mJy on the < 0.65 kl baseline of the RT, spatially coincident with the optical and X-ray positions for the cluster core. For a spherical King-profile cluster model, the best fit to our flux measurement has a core radius uC 20 arcsec, consistent with previous X-ray observations, and a central temperature decrement DT 650 ^ 92 mK. Using this model, we calculate that the cluster has a gas mass inside a 500 h6251 kpc radius of 2:9 1013 M( for an VM 1 universe and 1:6 1013 M( for VM 0:3, VL 0:7. We compare this model with existing measurements of the total mass of the cluster, based on gravitational lensing, and estimate a gas fraction for MS 1137.5 6625 of < 8 per cent. I N T R O D U C T I O N The cluster MS 1137.5 6625 was discovered in the Einstein Extended Medium Sensitivity Survey (EMSS; Gioia et al. 1990). It lies at redshift z 0:785 and has an X-ray temperature TX 6:0 keV and a 2 10 keV rest-frame luminosity of LX 2:8 1037 W (Gioia & Luppino 1994; Donahue et al. 1999). It is the second most distant cluster in EMSS, and is representative of the population of massive high-redshift clusters which is beginning to be discovered by X-ray selection and other means. Because the Sunyaev-Zeldovich (SZ; Sunyaev & Zeldovich 1972) effect is dependent on cluster gas mass, but is close to independent of redshift, we observed MS 1137 with the Ryle Telescope (RT) to attempt an SZ detection and constrain the cluster gas mass. RT observations totalling 460 h were made over 21 d in 1998 July and August and 18 d in 1999 March May. The telescope was in Cb configuration (Grainge et al. 1996), resulting in the aperture-plane coverage shown in Fig. 1. For each day, observations of the target field were interleaved with observations of a phase calibrator about every 20 min, and a primary flux calibrator (3C 286 or 3C 48) was observed at either the start or end of the run. The entire 39-d visibility data were concatenated and analysed to measure the cluster SZ signal. First, we made a map using only visibilities from projected baselines longer than 1.5 kl. First inspection of the dirty long-spacing map revealed two sources, the first lying close to the map centre, the second, about 30 arcmin to the south, is 3C 263 (Fig. 2). The long-spacing map was then CLEANed and the positions and fluxes of these sources measured with the task MAXFIT in the Astronomical Image Processing System (AIPS; http://www.cv.nrao.edu/aips/). Details of the sources are given in Table 1; The central source is clearly identified with the 2-mJy 20-cm source observed by Stocke et al. (1999). Two model point sources with the measured positions and fluxes were removed from all the visibilities using UVSUB in AIPS, and a new long-baseline map was made. This revealed only one further source close to the centre of the map with flux greater than 3.5 times the map rms. This source was removed from the visibility data using the same procedure. Then, a map was made using only the visibilities in the range 0:65 1:0 kl, corresponding to the shortest baseline. The dirty maps showed a clear negative feature at the centre, but also some residual flux from 3C 263. This map was CLEANed and the residual flux was measured and subtracted from the visibilities using MAXFIT and UVSUB. We then made a final short-spacing map in which the only feature is the SZ decrement of the cluster; the CLEANed version of this map is shown in Fig. 3. The decrement in the map has a minimum flux of 2422 ^ 60 mJy, a 7s detection (we take the noise to be the map RMS well outside the primary beam), and is centred at RA 11h 40m 20s:0, Dec. 668 070 5300 (J2000). Finally, we phase-rotated the source-subtracted visibilities to the decrement centre and azimuthally averaged them (Fig. 4). The mean flux on the shortest baseline is 2 421 mJy. The 1s positional error is roughly the beamwidth divided by the signal-tonoise ratio (e.g. Kenderdine, Ryle & Pooley 1966), i.e. about 20 arcsec. Thus the position of the SZ decrement is coincident with the optical and X-ray core of the cluster (e.g. fig. 2 of Donahue et al. 1999). The position and size of the decrement are also consistent with those measured by Grego et al. (2001) using the OVRO and BIMA arrays at 30 GHz. We next estimate the gas mass required to produce this SZ signal. We used PROFILE (Grainge et al. 2002) to model the cluster as a spherical King-profile gas distribution. Initially we used an VM 1 world model with H0 65 h65 km s21 Mpc21, taking the X-ray core radius uC 15 arcsec, TX 6 keV, and central electron density n0 1:6 104 h16=52 m23 values measured by Donahue et al. (1999). In this model, the SZ flux on the shortest RT spacing would be 2 339 mJy, 1.4 s different from our SZ measurement. The best-fitting model to the SZ data has n0 1:3 104 h16=52 m23 and uC 20 arcsec, and a central temperature decrement of DT 650 ^ 92 mK. For this model, we find a total gas mass enclosed inside a 500 h6251 kpc radius of 2:9 1013 M(; in an VM 0:3, VL 0:7 universe, the best-fitting model has n0 1:0 104 h16=52 m23 and uC 20 arcsec, giving a gas mass enclosed inside 500 h6251 kpc of 1:6 1013 M(. Clowe et al. (1998) have used a gravitational lensing analysis to estimate estimated the total mass of MS 1137, for an VM 1 universe. Comparing our best-fitting King-profile model with Clowe et al.s total mass estimate, we calculate a gas fraction inside a 500 h6251 kpc radius of 0:08 ^ 0:026. Finally, we note that the RT image shows essentially no substructure in MS 1137. This is consistent with the X-ray core radius measured by Donahue et al. (1999) and the compact mass distribution measured by Clowe et al. (1998). However, the distribution of galaxies in MS 1137, as noted by Clowe et al., has RA (J2000) 11 40 22.302 11 39 58.738 11 40 10.260 11 40 00.547 Dec. (J2000) 66 08 49.94 65 47 53.021 66 07 10.57 65 47 51.602 S15 GHz/mJy clear east west extensions; Clowe et al. propose that the apparent compactness of MS 1137 may be because we are observing several merging filaments, with one pointing along the line of sight. Unfortunately, the imaging capabilities of the RT are insufficient, in terms of both sensitivity and aperture coverage, to detect any gas associated with the proposed filaments. Targets such as MS 1137, where there may be gas filaments too faint for X-ray detection, will be ideal targets for next-generation SZ telescopes (e.g. Holder et al. 2000, Kneissl et al. 2001). 3 C O N C L U S I O N S (i) MS (...truncated)